E-Commerce Sales Analysis & Business Intelligence Dashboard

This project analyzes transactional e-commerce data to uncover revenue drivers, customer purchasing behavior, operational revenue leakage, and customer retention patterns.

The project combines:

• Python-based exploratory data analysis
• Data quality auditing
• Revenue intelligence analysis
• Customer segmentation
• Tableau dashboard development

The goal was to simulate a real-world business intelligence workflow rather than a basic visualization project.

Business Problem

The business needs to:

• Identify major revenue-driving products and customers
• Detect operational revenue leakage through returns
• Understand customer purchasing behavior
• Analyze customer concentration risk
• Improve retention and revenue strategy

Project Objectives

• Perform transaction-level data cleaning and auditing
• Build a transaction classification framework
• Analyze revenue concentration using Pareto analysis
• Segment customers based on purchasing behavior
• Identify operational return-risk products
• Build executive-level Tableau dashboards

Key KPIs Tracked

• Total Revenue
• Total Orders
• Average Order Value (AOV)
• Return Loss %
• Revenue Concentration %
• Customer Concentration %
• Repeat Customer Rate
• VIP Customer Revenue
• Operational Return Revenue Loss

Tech Stack

• Python
• Pandas
• NumPy
• Matplotlib
• Seaborn
• Tableau
• Jupyter Notebook

Dataset Information

Dataset Used: Online Retail Dataset (UCI Machine Learning Repository)

Dataset Characteristics:

• ~541K transactional records
• UK-based online retail transactions
• Data period: Dec 2010 – Dec 2011
• Includes invoices, products, quantities, pricing, customers, and countries

Dataset Features:

• Invoice-level transactional data
• Customer purchasing behavior
• Product-level revenue data
• Return/cancellation transactions
• International sales records

Dataset Challenges:

• Missing customer IDs
• Negative quantities (returns/cancellations)
• Financial adjustment transactions
• Duplicate records
• Zero-value operational transactions

Key KPIs Tracked

• Total Revenue
• Total Orders
• Average Order Value (AOV)
• Return Loss %
• Revenue Concentration %
• Customer Concentration %
• Repeat Customer Rate
• VIP Customer Revenue
• Operational Return Revenue Loss

Data Cleaning & Quality Audit

Key data quality issues identified:

• Missing Customer IDs (~25% of transactions)
• Negative quantities representing returns/cancellations
• Financial adjustment transactions
• Zero-value operational transactions
• Duplicate transaction rows
• Unspecified country values

A transaction classification framework was created to separate:

• Valid Sales
• Cancellation / Returns
• Financial Adjustments
• Non-Revenue Transactions
• Unknown Customer Transactions

Key Analyses

Revenue Intelligence

• Monthly revenue trends
• Revenue concentration analysis
• Product Pareto analysis
• Country-wise revenue analysis

Customer Intelligence

• Customer segmentation
• VIP customer analysis
• Repeat customer behavior
• Order frequency distribution
• Customer revenue concentration

Operational Intelligence

• Return revenue leakage analysis
• High-risk products
• Revenue vs operational risk analysis

Dashboard Previews

Executive Overview Dashboard

Product & Operational Intelligence Dashboard

Customer Intelligence Dashboard

Interactive Dashboard Features

• Country-level filtering
• Product-level operational risk exploration
• Customer segment drill-down analysis
• Interactive KPI storytelling
• Revenue concentration exploration
• Customer retention analysis

Interactive Tableau Dashboards

View Interactive Dashboards on Tableau Public:

Tableau Public Dashboard Link

Key Business Insights

• ~20% of products contributed nearly 80% of total revenue, revealing strong product concentration dependency.

• ~26% of customers generated 80% of total revenue, indicating significant reliance on high-value customer groups.

• VIP Loyal Customers contributed the majority of customer revenue, emphasizing the importance of customer retention strategies.

• Customer purchasing behavior showed a highly right-skewed distribution, with most customers placing only a small number of orders.

• Operational return revenue leakage was highly concentrated among a small subset of products.

• Returns and cancellations represented a meaningful source of revenue erosion requiring operational monitoring.

Strategic Recommendations

• Implement retention-focused loyalty strategies for VIP customers.
• Monitor high-return-risk products operationally.
• Reduce dependency on a small subset of revenue-driving products.
• Improve customer conversion from low-engagement to repeat purchasing behavior.
• Establish operational controls around return-heavy products.

How to Run

1. Clone the repository
2. Install dependencies
3. Open Jupyter Notebook
4. Run the analysis notebook
5. Open Tableau dashboards

Project Structure

ecommerce-sales-analysis/
│
├── data/
│   ├── raw/
│   └── cleaned/
│
├── notebooks/
│
├── dashboards/
│
├── images/
│
├── README.md
├── requirements.txt
└── .gitignore

Future Improvements

• Build SQL warehouse pipeline
• Develop Power BI version
• Add forecasting models
• Build customer lifetime value prediction
• Develop automated KPI reporting pipeline
• Deploy dashboards to Tableau Public
• Add RFM customer segmentation
• Build churn prediction models

Conclusion

This project demonstrates a business-oriented analytics workflow focused on transforming raw transactional data into strategic business intelligence.

The analysis combined:

• data cleaning
• operational auditing
• revenue intelligence
• customer analytics
• dashboard storytelling

to simulate a real-world analytics and business intelligence environment using Python and Tableau.